Publication detail

Cellular Neural Network in Noise Removing

LUDVÍK, J., VRBA, R.

Original Title

Cellular Neural Network in Noise Removing

English Title

Cellular Neural Network in Noise Removing

Type

conference paper

Language

en

Original Abstract

The paper discusses problems with a CNN CMOS implementation on an IC chip. Cell circuit is developed and simulated using HSpice simulator and CNN Java simulator. Final microelectronic structure is designed with these assumptions: simple structure, low number of transistors, small occupied area on the chip and high processing speed. The design is realised in Alcatel Mietec 0.35um CMOS technology. A functionality of the network for noise removing is proved for a network 3x3 cells and for a network of 8x8 cells, both networks exploiting the same neuone with fixed synaptic weights. The network of 8x8 cells implements added functions - the horizontal and vertical line detections.

English abstract

The paper discusses problems with a CNN CMOS implementation on an IC chip. Cell circuit is developed and simulated using HSpice simulator and CNN Java simulator. Final microelectronic structure is designed with these assumptions: simple structure, low number of transistors, small occupied area on the chip and high processing speed. The design is realised in Alcatel Mietec 0.35um CMOS technology. A functionality of the network for noise removing is proved for a network 3x3 cells and for a network of 8x8 cells, both networks exploiting the same neuone with fixed synaptic weights. The network of 8x8 cells implements added functions - the horizontal and vertical line detections.

Keywords

Neurone cell, IC chip, synaptic weight, noise removal, detection

RIV year

2002

Released

01.01.2002

Publisher

Vysoké učení technické v Brně

Location

Brno

ISBN

80-214-2180-0

Book

ELECTRONIC DEVICES AND SYSTEMS 02 - PROCEEDINGS

Edition number

1.

Pages from

330

Pages to

336

Pages count

7

BibTex 


@inproceedings{BUT4964,
  author="Jan {Ludvík} and Radimír {Vrba}",
  title="Cellular Neural Network in Noise Removing",
  annote="The paper discusses problems with a CNN CMOS implementation on an IC chip. Cell circuit is developed and simulated using HSpice simulator and CNN Java simulator. Final microelectronic structure is designed with these assumptions: simple structure, low number of transistors, small occupied area on the chip and high processing speed. The design is realised in Alcatel Mietec 0.35um CMOS technology. A functionality of the network for noise removing is proved for a network 3x3 cells and for a network of 8x8 cells, both networks exploiting the same neuone with fixed synaptic weights. The network of 8x8 cells implements added functions - the horizontal and vertical line detections.",
  address="Vysoké učení technické v Brně",
  booktitle="ELECTRONIC DEVICES AND SYSTEMS 02 - PROCEEDINGS",
  chapter="4964",
  institution="Vysoké učení technické v Brně",
  year="2002",
  month="january",
  pages="330",
  publisher="Vysoké učení technické v Brně",
  type="conference paper"
}